A dual mode FPGA implementation of real-time target detection for hyperspectral imagery

Target detection is an important topic for extracting information from hyperspectral imagery. Due to the high computational complexity, developing fast target detection algorithm has received considerable interest, specifically to take advantage of field-programmable gate array (FPGA) architecture in hardware implementation. In this paper, a novel streaming background statistics (SBS) method is used for optimizing the constrained energy minimization (CEM) and Reed-Xiaoli (RX) algorithms which are the benchmarks of signal-based target detection and anomaly detection, respectively. As a purpose, these two classic algorithms can be easily calculated in a streaming update way. Finally, we present a dual mode FPGA implementation in which, two different algorithms are arranged for different detection requirements within a real-time processing.